🍭 The Pope Just Picked His AI Lab ⛪

Good morning. The Pope published a 42,000-word manifesto on AI yesterday and literally cited Tolkien on how to resist dehumanization.

The AI safety discourse has lore now.

Let’s dive in 👇

🍭 What’s Cookin’:

• Pope Leo published his AI encyclical

• SpaceX S-1 reveals Anthropic compute cost

• Walmart's AI shopping agent is quietly printing money

Pope & AI
⛪ The Vatican's AI Manifesto

The Bite:

Pope Leo XIV published Magnifica Humanitas: a 42,300-word encyclical on AI and the first papal teaching document dedicated entirely to the technology.

The document was presented at the Vatican alongside Anthropic co-founder Christopher Olah, 135 years to the day after Pope Leo XIII published Rerum Novarum, the foundational Catholic text on workers' rights during the Industrial Revolution.

The encyclical covers AI-driven unemployment, autonomous weapons, data governance, transhumanism, and the role of education, repeatedly calling for slower adoption timelines and stronger regulation throughout.

Snacks:

• Signed May 15, released May 25. Timed exactly to the 135th anniversary of Rerum Novarum

• Calls mass AI unemployment a "grave evil" and "social calamity" requiring state intervention

• Final chapter dedicated entirely to AI in warfare, demanding "rigorous ethical constraints"

• Argues data should be treated as a common good, not left in private hands

• Quotes Tolkien's Return of the King on "small and steadfast acts of fidelity" against dehumanization

• Anthropic was the only AI lab invited to co-present. Not OpenAI, not Google, not Meta

Why it Bites:

The Vatican doesn't move fast, but when it does move, it moves for centuries.

Magnifica Humanitas isn't just a blog post or policy brief.

It's a teaching document that will shape how 1.4 billion Catholics think about AI for who knows how long.

And the stage management matters as much as the text.

And Anthropic being the only lab in the room?

That's the Vatican publicly endorsing one lab's safety posture over every other company building frontier AI.

OpenAI is filing an IPO.
Google just wrapped I/O.
Meta is spending $135 billion on AI infrastructure.

None of them got the call.

For Anthropic, this is a legitimacy asset that money can't buy: a moral institution with a billion-person audience saying, in effect, "these are the people we trust to be in the room."

For everyone else building AI, the encyclical is a 42,000-word reminder that the people asking you to slow down aren't going away.

I mean, they have The Pope now.

Plaid
📈 AI is raising the bar in finance

Over half of Americans have used AI to manage their finances.
And they're not going back. 

The State of Intelligent Finance, Plaid's latest report in partnership with the Harris Poll, surveyed 2,000+ American adults on how AI is changing what people expect from financial products.

Consumers want products that understand their situation, guide their decisions, and act on their behalf. Trust hinges on transparency and accountability.

The report covers:

• Where consumers are most ready to let AI take action?

• What builds trust and what breaks it?

• How teams can turn rising expectations into a competitive advantage?

Steal This Prompt
🎋 Japanese Ink Wash, But Make It Existential

Turn literally anything into a moody Japanese ink wash painting that looks like it belongs in a Kyoto tea house and/or an A24 fever dream.

Use it to:

• Turn portraits into Zen-master-core profile pics

• Make products look weirdly expensive

• Transform pets into ancient mythical beings

Workflow:

1. Hit this link: Japanese Ink Wash Painting

2. Paste into your AI model

3. Replace the #s with your subject and colors

4. Watch it cook: suddenly your dog looks like he guards sacred mountains

   Try the prompt on Snack Prompt

_ToolBox™
🧰 5 BRAND NEW AI LAUNCHES

🚀 MiniCPM5-1B
A tiny open model with main-character energy. 131K context, tool calling, GGUF + MLX support, and built to run locally without setting your laptop on fire.

🧠 Hayley
Talk for 60 seconds. No prompts. No structure. Just raw brain static. Then it reflects the patterns you keep looping on like an emotionally intelligent mirror.

📬 ReplylessAI Sequences
Follow-up emails without buying enterprise software that feels like it was designed by a committee in 2009.

🎭 AVTR-1 Real-Time Open Weights Model
Real-time AI avatars that react while you speak. No creepy animation loops. Just generated faces vibing in sync with your voice.

🖥️ DNSimple CLI
DNS and domain management for terminal enthusiasts. Register domains, manage SSLs, and update records without touching a browser tab.

_{Everything Else}
🧠 You Need to Know

🚀 SpaceX S-1 Reveals How Much Anthropic Pays For Compute
→ The IPO filing disclosed that the Colossus deal is worth up to $45B through May 2029, with either party able to exit on 90 days' notice.

🧠 Karpathy Joins Anthropic's Pre-Training Team
→ The OpenAI co-founder and former Tesla AI lead started this week under Nick Joseph, building a team focused on using Claude to accelerate pre-training research.

🛒 Walmart Says Sparky AI Users Spend 35% More
→ CEO John Furner reported Sparky weekly active users doubled last quarter, with the AI shopping agent now available in stores and in Spanish.

⚙️ Anthropic Acquires Stainless For $300M+
→ The startup that generated every official SDK for Anthropic, OpenAI, and Google is now Anthropic-exclusive, with hosted products being shut down for all other customers.

⛪ Pope Leo Publishes First AI Encyclical
→ The 42,300-word Magnifica Humanitas, released alongside Anthropic co-founder, frames unchecked AI development as a modern Tower of Babel and calls for rigorous ethical constraints on autonomous weapons.